1. health and fitness
2. disease
3. asthma

Diffusing Capacity: How to Get It Right
Robert L Jensen PhD and Robert O Crapo MD
Introduction
Calculating Diffusing Capacity
Measuring Diffusing Capacity
Calibrating the Equipment
Optimizing the Test Procedure
Summary
The carbon monoxide diffusing capacity test (DLCO) is a commonly performed pulmonary function test
that requires technical expertise and attention to detail to get acceptable results. With the advent of
automated devices and powerful computer programs, DLCO measurement has rapidly gained wide
clinical acceptance. But there are many subtle aspects to performing the test that can diminish its
accuracy and repeatability. The clinician must ensure: that the DLCO instrument is correctly calibrated;
that inhalation is least 90% of the largest previously measured vital capacity; that the patient executes
a quick, smooth inhalation within 2 seconds; that the breath-hold is 9 –11 seconds; that the breath-hold
is without straining (no Valsalva or Mu¨ller maneuvers); that exhalation is quick and smooth; that a
representative gas sample is obtained from the correct portion of the exhalation; and that at least 5
minutes elapse between DLCO tests. At least 2 but no more than 5 DLCO tests should be conducted, and
testing is complete when 2 tests are within 10% or 3 DLCO units (mL CO/min/mm Hg) of each other. The
reported DLCO value is the average of the first 2 tests that meet the reproducibility criteria, but if 5 tests
are performed and no 2 meet the reproducibility criteria, the reported value is the average of the 2 tests
with the highest inspiratory volumes. These quality controls will help laboratories achieve consistent
high DLCO accuracy. Key words: diffusing capacity, DLCO, carbon monoxide, pulmonary function tests. [Respir
Care 2003;48(8):777–782. © 2003 Daedalus Enterprises]
Introduction
The efficiency of the lung in transporting oxygen from
the alveoli across the alveolar-capillary membrane and into
hemoglobin within the red blood cells can be assessed by
Robert L Jensen PhD and Robert O Crapo MD are affiliated with the Pulmonary
Division, LDS Hospital, and the University of Utah, Salt Lake City, Utah.
Dr Jensen presented a version of this report at the 18th Annual New
Horizons Symposium, Pulmonary Function Testing in 2002: Updates and
Answers, October 6, 2002, at the American Association for Respiratory
Care’s 48th Internaitonal Respiratory Congress in Tampa, Florida.
Correspondence: Robert L Jensen PhD, Pulmonary Division, LDS Hospital, 8th Avenue and C Street, Salt Lake City UT 84143. E-mail:
ldrjens1@ihc.com.
RESPIRATORY CARE • AUGUST 2003 VOL 48 NO 8
the carbon monoxide diffusing capacity test (DLCO).1,2 This
is because carbon monoxide (CO) follows the same diffusion pathway as oxygen, and its uptake is relatively easy
to measure. Like oxygen molecules, CO molecules move
from the alveolar space across the lung membrane, into the
plasma, into the red blood cells, and finally bind onto the
hemoglobin molecules at the same sites as do oxygen molecules (Figure 1). The only major differences between CO
and oxygen behavior are (1) the solubility of the gases in
intercellular fluids and plasma and (2) how strongly these
gas molecules adhere to the hemoglobin binding sites.3
Calculating Diffusing Capacity
The single-breath DLCO test has been made widely available by manufacturers that have automated and comput-
777
DIFFUSING CAPACITY: HOW
Fig. 1. The gas diffusion pathway is similar for carbon monoxide
(CO) and oxygen (O2). Hb ⫽ hemoglobin.
erized the procedure. All commercial systems conduct the
test during a single-breath maneuver. A tracer gas (helium,
neon, or methane) is also used to measure the initial dilution of the inhaled CO and to estimate alveolar volume.
The fundamental equation to calculate DLCO is:
DLCO ⫽
˙ CO
V
PACO ⫺ Pc៮ CO
(1)
˙ CO is the rate of disappearance of CO; PACO is
in which V
the alveolar concentration of CO, which is determined
from measuring the exhaled gas after the anatomical dead
space (trachea and upper airways) has been cleared; and
Pc៮ CO is the partial pressure of CO in the blood.
We can assume Pc៮ CO is essentially zero, except in smokers and individuals exposed to CO. With this assumption
the final equation then becomes:
DLCO ⫽
˙ CO
V
PACO
(2)
In North America the unit for DLCO is mL CO/min/mm Hg.
The Syste`me Internationale unit is mmol/min/kPa, and in
Europe the test is called the CO transfer factor. The following
equation converts between the 2 units:
DLCO ⫽ 2.986 ⫻ TLCO(SI)
(3)
Measuring Diffusing Capacity
The basic measurement procedure is straightforward.
The subject inhales a test gas containing 0.3% CO and a
tracer (usually 0.3% CH4, 0.5% Ne, or 1–5% He), and a
measurement is made of the exhaled concentrations of CO
and the tracer after a breath-hold of about 10 seconds.
The single-breath DLCO is not the only way to measure
CO transport across the lung membrane. The intra-breath
778
TO
GET IT RIGHT
method, the steady state, and rebreathing methods can also
measure CO transport. However, the single-breath DLCO
has been extensively studied within healthy subjects, reference equations have been derived, and automated devices are commercially available; therefore, the singlebreath DLCO is at present the only method finding clinical
utility. Further development is required before other DLCO
methods will have similar clinical acceptance.
The American Thoracic Society (ATS) has adopted standardized testing procedures and equipment recommendations for the single-breath DLCO.4 Standardization of testing procedures and equipment ensures that the test is
performed consistently. These standards allow comparison
of measurements from different pulmonary function laboratories. Several small studies have found that differences
between measurements made with the same individual in
several different laboratories can vary as much as 50%.5
Most the differences can be explained by procedural problems that, when addressed, reduce the differences.
In measuring DLCO the first thing to recognize is that
there are interactions: (1) between the subject and the testing device (eg, equipment placed in front of the patient’s
face, mouthpieces), (2) between the technician and the
subject, and (3) between the technician and the testing
device. The condition of the subject is critical to the test.
Several patient factors can affect the results. The subject
should not have had any of the following: a meal within 2
hours of the test; recent strenuous exercise; ongoing or
recent respiratory infection; or a blood alcohol level. The
subject should already have had spirometry (to measure
forced vital capacity or slow vital capacity), should be
fully cooperative and able to follow instructions, and should
be seated for the test session.
Clear explanation of the testing procedure prior to the
test is essential to obtaining good results. Demonstrating
the maneuver to the patient, with a mouthpiece, can overcome some training and language problems.4
The standard procedure is to test the patient while he or
she is seated in a stationary chair. The test should always
be done with nose clips and the mouthpiece and valves
placed at a comfortable level and position for the patient.4
Most devices have a screen display of patient breathing
patterns prior to performing the test. During the “prebreath-hold” phase the patient is instructed to exhale completely (though this is not a forced maneuver), at which
point valves are activated and the patient is instructed to
inhale the test gas fully, in ⱕ 2 seconds, and hold the
breath. It is very important that the patient rest against a
closed glottis or the valve and not strain to pull in (Mu¨ller
maneuver) or apply pressure by bearing down (Valsalva
maneuver). The Mu¨ller maneuver will recruit blood to the
lung and cause an inaccurately high DLCO reading, and the
Valsalva maneuver will decrease the volume of blood in
the lung and cause an inaccurately low DLCO reading. The
RESPIRATORY CARE • AUGUST 2003 VOL 48 NO 8
DIFFUSING CAPACITY: HOW
breath-hold period is about 10 seconds: the ATS standard
is 9 –11 seconds.4 Most DLCO measuring devices display
or indicate the end of the breath-hold period, and the patient is coached to exhale fully when the valves open or the
time interval is completed. The ATS standard waiting period between consecutive DLCO maneuvers is 5 min, which
ensures that all the test gas has washed out of the lungs.
Even though the test seems simple, the results are affected by several factors, the most important of which are:
1. Inspired volume
2. Calibration
3. Sufficient washout volume
4. Representativeness of the sampled gas
The inspired volume is critical to the test. Imagine if a
subject only inhaled enough test gas to fill the anatomical
dead space. No CO diffusion would occur and the effective DLCO measurement would be zero, since CO can only
diffuse into the blood when it reaches the alveolar region.
Even if the test gas enters the alveolar region, if the lung
is not fully inflated, CO uptake will be inappropriately
reduced because of incomplete membrane expansion. Thus,
the inhalation must be at least 90% of the patient’s largest
measured vital capacity, to ensure sufficient inhaled volume. This is why it is necessary to measure the patient’s
vital capacity prior to the DLCO test. If the inspiratory
volume is ⬍ 90% of the vital capacity the DLCO measurement will be inaccurately low and the patient might be
misinterpreted as having below-normal DLCO.
The way to obtain a good vital capacity measurement is
the same as for all good quality spirometry. Obtain the
following for each spirometry effort:
1. Full inspiration to total lung capacity
2. Quick forced expiration, without hesitation
3. Minimum of 6 seconds of exhalation
4. At least 3 acceptable trials
5. The reproducibility of the 2 best tests should be ⫾
200 mL for both forced vital capacity and forced expiratory volume in the first second
There are many excellent guides and books about spirometry testing. Every effort should be made to meet the
ATS standards in the ATS manual of procedures.6 Though
no laboratory can have all patients meet the ATS standards, even in laboratories that see numerous very sick
patients the ATS standards can be met ⬎ 85% of the time.
Calibrating the Equipment
Calibration is always critical. It is the essential element
in good quality control and accurate measurements. The
ATS standards recommend that DLCO systems be calibrated daily, and twice daily in busy laboratories. Fortunately, the manufacturers have almost uniformly instrumented DLCO systems to perform gas calibrations just prior
to each DLCO test. This reduces problems associated with
RESPIRATORY CARE • AUGUST 2003 VOL 48 NO 8
TO
GET IT RIGHT
normal drift in the gas analyzer outputs, which greatly
influence test results. However, the flow sensors and/or
volume sensors are not recalibrated prior to each DLCO test
and need to be recalibrated daily or twice daily. The volume measurement systems should be calibrated using a
3-L syringe that is in good condition and has had its accuracy certified by the factory or some other reputable
source at some regular interval (1 or 2 y).
Barometric pressure and temperature are also important.
The DLCO calculations use barometric pressure measurements, so they should be accurate on average. The barometric pressure should be set to the normal average barometric pressure at your laboratory. Weather-related changes
in barometric pressure will not greatly affect DLCO measurements. Temperature on the other hand is very important. Each 1° C of mismeasurement introduces a 0.67%
error in the DLCO. That may not seem like much, but
normal temperature variations in a laboratory can be up to
10° C per day and could introduce up to 7% error in the
DLCO if not included in the DLCO calculations. Use a good
electronic thermometer and consistently enter the temperature into the DLCO instrument to increase your DLCO accuracy.
Confidence in volume measurements allows you to make
correct judgments about the adequacy of inspired volumes.
If these systems are not correctly volume-calibrated, a
number of problems may occur:
1. The inspiratory volume can be erroneous, which can
lead to incorrectly accepting or rejecting a maneuver based
on the 90%-of-vital-capacity criteria
2. If inspiratory volume is mismeasured, then alveolar
volume will be incorrect
3. DLCO may be incorrectly calculated
4. Valves may not open or close at the correct volumes
or times
One problem occasionally encountered is that the volume calibration for the DLCO system is done when the
system is in a spirometry mode or configuration, after
which the system is changed to a DLCO mode that may
require adding valves or tubing to be placed near the flowsensor. Those changes in the physical configuration can
alter the volume calibration characteristics.
A simple check of the DLCO volume calibration can be
done with a 3-L syringe DLCO test. First set up the system
to do the test for a normal patient. Next attach a 3-L
syringe to the mouthpiece. When the syringe is completely
emptied, start the DLCO inhalation maneuver. Withdraw
the full 3 L into the syringe, wait approximately 10 seconds, and then discharge the syringe back into the device.
The measured inspiratory volume should be 3 L multiplied
by the correction factor converting ambient-temperatureand-pressure-dry (ATPD) to body-temperature-and-pressure-saturated (BTPS). This factor is typically around 1.10,
making the measured inspiratory volume around 3.33 L.
779
DIFFUSING CAPACITY: HOW
TO
GET IT RIGHT
Fig. 3. Ideal tracing of a single-breath DLCO test. In the initial phase
the inhaled volume and gas concentrations rise. The plateau of the
curve represents the breath-hold phase. In the exhalation phase
lung volume decreases sharply (%VC ⫽ percent of vital capacity),
whereas carbon monoxide concentration ([CO]) decreases more
gradually, and the tracer decreases and then reaches a plateau.
Fig. 2. A DLCO simulator designed to test the accuracy of DLCO
measurement instruments. The simulator produces known volumes of precision gases that simulate human breath. The device
can reliably simulate the same value to within fractions of a DLCO
unit (mm CO/min/mm Hg). (Courtesy Hans Rudolph Inc, Kansas
City, Missouri)
The measured DLCO should be zero, since no CO diffusion
can occur in the syringe.
A new device has become available for testing DLCO
instruments (Figure 2).7 It uses precision syringes in conjunction with precision mixed gases to simulate volumes
and gas concentrations that would be measured from a
human subject. A software calculator records the values
the instrument measures and reports differences for each
aspect of the DLCO measurement, including gas concentrations, volumes, and final DLCO calculations. If the DLCO
measurement device does not report the target DLCO, the
problem can be easily identified as a volume, carbon monoxide, or tracer measurement failure.
Optimizing the Test Procedure
The washout volume is the volume of gas that must
leave the lungs so that all remaining gas exiting the lungs
is gas from the alveolar gas-exchanging regions and not
780
from the trachea and upper airways, where no gas exchange occurs. The ATS standards recommend a minimum washout volume of 750 mL.4 Some instruments use
rapid gas analyzers that display the gas concentration curves
and allow the user to adjust the washout volume. For those
types of instruments the washout volume is adjusted so the
tracer concentration reaches a plateau and the CO concentration indicates alveolar gas (Figure 3).
Obtaining a representative sample of the alveolar gas is
key to obtaining a representative DLCO measurement. When
rapid gas analyzers are used and the graphical output of
these is displayed for review, determining if the gas samples are from the correct part of the exhaled gas volume is
made simple by visually inspecting the curves. Somewhere
between 0.75 and 1.5 L of gas should be used for the gas
analysis. When curves are not available, we can only assume that the washout and sample volumes are appropriate
and that the correct gas concentration measurements were
made (Figure 4).
Achieving good control of the technical aspects of the
DLCO test is the first step toward getting excellent results.
Applying quality control procedures to your laboratory will
help ensure that your results are consistently excellent.
Quality control for the DLCO test includes (1) immediate
appraisal of the test and (2) reproducibility of multiple
test maneuvers. Immediate appraisal includes:
1. Inhalation to at least 90% of the largest previously
measured vital capacity
2. Quick, smooth inhalation within 2 seconds
3. Breath-hold of 9 –11 seconds
4. Hold the breath without straining (no Valsalva or
Mu¨ ller maneuvers)
5. Quick smooth exhalation
6. Check the washout and sample volumes if the device
allows
7. Wait at least 5 min between DLCO tests
RESPIRATORY CARE • AUGUST 2003 VOL 48 NO 8
DIFFUSING CAPACITY: HOW
TO
GET IT RIGHT
Fig. 4. An actual DLCO tracing, showing the instrument’s automated selection for the washout volume (discard) and the sample volume
(collection). The upper panel shows the first trial and lower panel shows the second trial. Note the differences in washout volumes
(Trial 1 ⫽ 1.21 L, Trial 2 ⫽1.66 L) and collection volumes (Trial 1 ⫽ 0.87 L, Trial 2 ⫽ 0.75 L); these values are determined from the
part of the curve that decreases after the plateau. Correct setting of the washout and collection volumes will produce different values
for each test.
RESPIRATORY CARE • AUGUST 2003 VOL 48 NO 8
781
DIFFUSING CAPACITY: HOW
These quality control measures ensure that each individual test is performed well. Following these guidelines
can help identify methodological problems that might lead
to erroneous DLCO values. When problems are identified,
an individual test can be discarded rather than sent for
interpretation.
Reproducibility is the second aspect:
1. Perform at least 2 but no more than 5 DLCO tests
2. Testing is complete when 2 tests are within 10% or 3
DLCO units (mL CO/min/mm Hg) of each other
3. The reported DLCO value is the average of the first 2
tests that meet the reproducibility criteria. However, if 5
tests are performed and no two meet the reproducibility
criteria, the reported value is the average of the 2 tests with
the highest inspiratory volumes.
Following these simple quality control procedures will
reduce noise in your data, ensure that your DLCO values
are accurate, allow better diagnosis with the DLCO values,
and allow better assessment of DLCO trends or changes
when patients return for follow-up tests.
The choice of reference values for DLCO is made by the
individual laboratory.8 However, several surveys have
found that a large number of laboratories were using the
reference values that were the factory defaults when they
purchased the device. Each laboratory’s staff should review with their medical director their choice of reference
equations and should document their choice. See Crapo
and Jensen’s review of reference equations in this issue of
RESPIRATORY CARE.9
There is no way to describe the complex understanding
that each pulmonary technician has of the equipment, laboratory, and individual characteristics of each patient being tested. To obtain consistently representative results, be
alert for peculiarities in the testing, abnormally high or low
782
TO
GET IT RIGHT
gas concentrations, or suspicious attributes or breathing
patterns, and always consider if the results seem to fit the
clinical impressions for the subject.
Summary
The clinician must pay careful attention to calibration
and the test procedure to maximize the consistency and
value of DLCO results.
REFERENCES
1. Comroe JH Jr. Pulmonary diffusing capacity for carbon monoxide
(DLCO). Am Rev Repir Dis 1975;111(2):225–228.
2. Krogh M. The diffusion of gases through the lungs of man. J Physiol
1914;49:271–300.
3. Forster RE, Fowler D, Bates V, Van Lingen B. The absorption of
carbon monoxide by the lungs during breathholding. J Clin Invest
1954;33:1135–1145.
4. American Thoracic Society. Single-breath carbon monoxide diffusing capacity (transfer factor): recommendations for a standard technique—1995 update. Am J Respir Crit Care Med 1995;152(6 Pt
1):2185–2198.
5. Wanger J, Irvin C. Comparability of pulmonary function results from
13 laboratories in a metropolitan area. Respir Care 1991;36(12):
1375–1382.
6. Wanger J. Editor-in Chief. Pulmonary function laboratory management and procedure manual. A project of the American Thoracic
Society. 1998. Available at http://www.thoracic.org/education/labmanual.asp (accessed 6/11/03).
7. Glissmeyer EW, Jensen RL, Crapo RO, Greenway LW. Initial testing with a carbon monoxide diffusing capacity simulator (abstract).
J Investig Med 1999;47:37A.
8. American Thoracic Society. Lung function testing: selection of reference values and interpretative strategies. Am Rev Respir Dis 1991;
144(5):1202–1218.
9. Crapo RO, Jensen RL. Standards and interpretive issues in lung
function testing. Respir Care 2003;48(8):764–772.
RESPIRATORY CARE • AUGUST 2003 VOL 48 NO 8